Multiplex PCR−Based Next-Generation Sequencing and Global Diversity of Seoul Virus in Humans and Rats

Seoul virus (SEOV) poses a worldwide public health threat. This virus, which is harbored by Rattus norvegicus and R. rattus rats, is the causative agent of hemorrhagic fever with renal syndrome (HFRS) in humans, which has been reported in Asia, Europe, the Americas, and Africa. Defining SEOV genome sequences plays a critical role in development of preventive and therapeutic strategies against the unique worldwide hantavirus. We applied multiplex PCR–based next-generation sequencing to obtain SEOV genome sequences from clinical and reservoir host specimens. Epidemiologic surveillance of R. norvegicus rats in South Korea during 2000–2016 demonstrated that the serologic prevalence of enzootic SEOV infections was not significant on the basis of sex, weight (age), and season. Viral loads of SEOV in rats showed wide dissemination in tissues and dynamic circulation among populations. Phylogenetic analyses showed the global diversity of SEOV and possible genomic configuration of genetic exchanges.

virus genome sequences. However, an obstacle for obtaining virus genomic information is ultra-low virus RNA loads in the clinical specimens. To enrich the low amount of viral RNA, we developed a multiplex PCR-based NGS that showed high coverage of HTNV genome sequences from HFRS patients (19).
In this study, we collected 1,269 R. norvegicus rats in an urban HFRS-endemic area in South Korea during 2000-2016. We report a robust strategy for whole-genome sequencing of SEOV and provide useful insights into epidemiologic characteristics and phylogeographic diversity of a unique worldwide hantavirus.

Ethics
Human samples were provided after informed consent was obtained. The study was approved and conducted in accordance with ethics guidelines for the Korea University Institutional Animal Care and Use Committee. Live trapping of rats at US military training sites and installations was approved by US Forces Korea in accordance with regulation 40-1 (Prevention, Surveillance, and Treatment of Hemorrhagic Fever with Renal Syndrome). Rats were humanely killed by cardiac puncture, and tissues were collected under isoflurane anesthesia in accordance with procedures approved by Korea University Institutional Animal Care and Use Committee protocol #2010-212.

Sample Collection
We tested retrospective HFRS patient serum samples obtained from the Korea Bank for Pathogenic Viruses (Seoul, South Korea). We collected R. norvegicus rats during 2000-2016 by using collapsible live-capture traps (Tomahawk Live Trap Co., Hazelhurst, WI, USA, and H.B. Sherman, Tallahassee, FL, USA). Traps were set at intervals of 1-2 m and examined early the next morning over a 1-2-day period at US Army training sites. For the US Army Garrison in Seoul, we used baited live capture traps (Tomahawk Live Trap Co.) or glue boards. Captured rats were submitted to the 5th Medical Detachment/Medical Command Activity-Korea, US Army Garrison (Yongsan, Seoul), and then transported to the College of Medicine, Korea University (Seoul), where they were held in a Biosafety Level 3 laboratory until processing. Live rats were humanely killed by cardiac puncture under isoflurane anesthesia and identified to species by using morphologic criteria and PCR, when required. Serum, lung, spleen, kidney, and liver tissues were collected aseptically and frozen at -70°C until used.

Indirect Immunofluorescence Antibody Test
We used an indirect immunofluorescence antibody (IFA) test for serum samples from HFRS patients and live rats.
We initially diluted samples 1:32 in phosphate-buffered saline and then tested them for IgG against SEOV. We applied diluted serum samples to slides containing SEOV-infected Vero E6 cells fixed with acetone and incubated wells at 37°C for 30 min. The slides were washed, fluorescein isothiocyanate-conjugated goat antibody to human and rat IgG (ICN Pharmaceuticals, Laval, Quebec, Canada) was added, and slides were incubated at 37°C for 30 min. We then washed the slides again and examined them for virusspecific fluorescence by using a fluorescent microscope (Axio Scope; Zeiss, Berlin, Germany).

Real-Time Quantitative PCR
We performed real-time quantitative PCR (qPCR) for total RNA by using the high-capacity RNA-to-cDNA Kit (Applied Biosystems, Carlsbad, CA, USA) in a 10-µL reaction mixture containing 1 µg of total RNA. We used an SYBR Green PCR Master Mix (Applied Biosystems) in a StepOne Real-Time PCR System (Applied Biosystems). We performed reactions at 95°C for 10 min, followed by 45 cycles at 95°C for 15 s, and then 1 cycle at 60°C for 1 min. Primer sequences specific for SEOV S segments were SEOV-S719F: 5′-TGGCACTAGCAAAAGACTGG-3′ and SEOV-S814R: 5′-CAGATAAACTCCCAGCAATAGGA-3′.

Reverse Transcription and Rapid Amplification of cDNA Ends PCR
We extracted total RNA from serum or lung tissues of seropositive samples by using TRI Reagent Solution (Ambion Inc., Austin, TX, USA). We synthesized cDNA by using the High Capacity RNA-to-cDNA Kit (Applied Biosystems) and random hexamer or OSM55 (5′-TAG-TAGTAGACTCC-3′). For initial identification, we used oligonucleotide primers for SEOV L segment as described (20). To obtain the 3′ and 5′ termini genome sequences of SEOV, we performed rapid amplification of cDNA ends (RACE) PCR by using a 3′-Full RACE Core Set and a 5′-Full RACE Core Set (Takara Bio Inc., Kusatsu, Shiga, Japan), according to the manufacturer's specifications. We purified PCR products by using the LaboPass PCR Purification Kit (Cosmo Genetech, Seoul, South Korea). We performed sequencing in both directions of each PCR product by using the BigDye Terminator v3.1 Cycle Sequencing Kit (Applied Biosystems) on an automated sequencer (Applied Biosystems).
We performed the first and second enrichments in a 25-µL reaction mixture containing 12.5 µL of 2× Uh pre-mix, 1 µL of cDNA template, 10 µL of primer mixture, and 1.5 µL of distilled water. Initial denaturation was at 95°C for 15 min, followed by 40 cycles or 25 cycles at 95°C for 20 s, 50°C for 40 s, and 72°C for 1 min, and a final elongation at 72°C for 3 min.
We prepared multiplex PCR products by using the TruSeq Nano DNA LT Sample Preparation Kit (Illumina, San Diego, CA, USA) according to the manufacturer's instructions. We mechanically sheared samples by using an M220 focused ultrasonicator (Covaris, Woburn, MA, USA). The cDNA amplicon was size-selected, A-tailed, ligated with indexes and adaptors, and enriched. We sequenced libraries by using the MiSeq benchtop sequencer (Illumina) with 2 × 150 bp and a MiSeq reagent V2 (Illumina). We imported and analyzed Illumina FASTQ files by using EDGE (21).

Phylogenetic Analysis
We aligned and edited virus genome sequences by using the multiple sequence alignment with high accuracy and high throughput algorithm (22). We generated phylogenetic trees by using the maximum-likelihood method in MEGA version 6.0 (23) and models for analysis according to the best fit substitution model (TN93 + gamma + invariate for L segments, general time reversible + gamma + invariant for M segments, and T92 + gamma for S segments). We assessed support for topologies by bootstrapping for 1,000 iterations. The prototype strain used, SEOV 80-39, was isolated from R. norvegicus rats captured in Seoul in 1980.

Retrospective Analysis of HFRS Patient Specimens
We found that specimens collected in 2002 from 6 HFRS patients were positive for SEOV by ELISA (J.-W. Song and H. Kariwa, unpub. data). We confirmed that the HFRS specimens were serologically positive for SEOV by IFA ( Table 1). Titers of SEOV-specific antibody ranged from 1:128 to 1:4,096. Reverse transcription PCR detected the partial sequence of L segment (nt 2946-3335) from 2 HFRS patients (Hu02-180 and Hu02-258).

R. norvegicus Rats
To measure viral load of SEOV RNA in R. norvegicus rats, we performed real-time qPCR for seropositive samples from lungs, livers, kidneys, and spleens ( Figure 1). Viral load for SEOV showed ranges from tissues of 5 rats (Rn02-15, Rn10-134, Rn10-145, Rn11-44, and Rn11-53) that were positive by serologic and molecular screening (IFA+ PCR+). Rat Rn10-145 showed the highest amount of SEOV RNA in all tissues, followed by rats Rn02-15 and Rn10-134. Rat Rn11-44 showed the highest amount of SEOV RNA in all tissues except liver. Rat Rn11-53 showed the highest amount of SEOV RNA load in lung tissues, but virus RNA was not detectable in liver, kidney, and spleen tissues.

Multiplex PCR-Based NGS for Retrospective HFRS Patient and R. norvegicus Rat Specimens
We determined viral loads for HFRS patient specimens by using real-time qPCR. Cycle threshold (C t ) values ranged from 27.5 to 36.8 (Table 3). To perform multiplex PCR-based NGS for SEOV, we designed multiplex PCR primers to amplify  We obtained 9 genome sequences of SEOV S segments from HFRS patients and R. norvegicus rats. Phylogenetic analysis of SEOV S segments showed that group A formed a monophyletic lineage with group D (Figure 2). Group C genetically clustered with group E. The phylogeny of group B was distinct from those of groups A, C, D, and E. Group F from mountainous areas in China formed a lineage that was independent from the other groups obtained from rats collected in urban areas.
We obtained 6 genome sequences of SEOV M segments from HFRS patients and R. norvegicus rats ( Figure  3). Phylogenetic analysis of SEOV M segments showed distinct phylogenetic clusters (groups A-F). These phylogenetic patterns showed that M segments of SEOV had genetic heterogeneity when compared with S segments.
We obtained and phylogenetically analyzed 5 SEOV L segments (Figure 4). The SEOV L segment from an HFRS patient and R. norvegicus rats captured in South Korea clustered to form a monophyletic group with SEOV 80-39. SEOV strains from China belonged to a genetic lineage with SEOV DPRK08 from North Korea. SEOV strains from the United Kingdom and Baltimore formed a close phylogenetic group. SEOV IR33 and IR473 obtained from laboratory outbreaks in the United Kingdom were independent from other SEOV strains.

Discussion
NGS is a robust tool for obtaining extensive genetic information and completing whole-genome sequences (18). However, molecular enrichment plays a critical role in amplification of pathogen genomic sequences from clinical or animal specimens. Our previous study showed recovery of nearly whole-genome sequences of HTNV from HFRS military patients by using virus-targeted molecular enrichment (19).
In this study, whole-genomic sequencing of SEOV, an etiologic agent of mild HFRS worldwide, was applied to samples from retrospective HFRS patients and seropositive R. norvegicus rats by using multiplex PCR-based NGS. Nearly whole-genome sequences of SEOV tripartite RNA, on the basis of SEOV 80-39 (prototype strain), corresponded to viral loads of patient serum samples and rat lung tissues. Phylogenetic analyses of the genome sequence of SEOV tripartite RNA supported worldwide distributions of SEOV and identified 6 genetic lineage groups. Group A contained SEOV strains from northeastern and southeastern China and North Korea. Group B contained SEOV strains from Singapore and Vietnam in Southeast Asia and Lyon in France. Group C contained SEOV strains originating primarily in South Korea and Japan and an SEOV strain from Louisiana in the United States. Group D consisted of SEOV strains from southeastern China, including Jiangxi and Hubei Provinces. Group E contained SEOV strains from the United Kingdom and eastern United States (New York and Baltimore) and formed a monophyletic lineage. Group F contained SEOV strains from mountainous areas in southeastern China (24).
SEOV originated in China and spread worldwide during movement of rats coincidently with human activities (e.g., commercial trade, travel, and migration by railways and through seaports) (24). The close genetic relationship of SEOV in South Korea and Japan was probably caused by geographic distance and historical activities (e.g., commerce and occupation by Japanese forces). The genetic lineage containing strains from Southeast Asia and France might have originated during colonization or on trade routes that extended distribution of SEOV-infected rats (25). Recently, SEOV outbreaks have been reported in the United Kingdom and United States. Clinical cases showed that SEOV infections were identified among pet owners, breeders, and distributors (26). The genetic relationship of SEOV between counties probably reflects movement of rats associated with the animal pet market. The prevalence of hantaviruses (e.g., HTNV and Imjin virus [MJNV]) in natural reservoir hosts has showed sexand weight (age)-specific differences (27,28). However, in our study, the incidence of SEOV in R. norvegicus rats was not dependent on sex and weight (age). Epidemiologic differences in hantavirus infections between A. agrarius and R. norvegicus rats might be, in part, caused by ecologic differences, reservoir host distributions, and behavior (e.g., association with humans) (29) an enzootic infectious cycle. These observations might suggest that preventive strategies for disease risk mitigation focus on limits of rat populations all year.
Our previous study demonstrated differential amounts of HTNV RNA in lung, kidney, liver, and spleen tissues of rodents collected in areas in which HFRS is prevalent (30). In addition, the genomic RNA load of MJNV, a shrewborne hantavirus, showed various patterns in different tissues in nature (28). IFA+ PCR+ shrews showed high and various loads of MJNV RNA in all tissues. MJNV RNA from IFA-PCR+ shrews was detected in lung but not in kidney, liver, or spleen tissues, indicating an early phase of infection before MJNV-specific IgG was produced (31,32). In our study, rats Diversity of virus genomes results from genomic variation or exchanges (33). RNA viruses show high mutation rates caused by deficiencies in proofreading by virus polymerases. Genomic variation also results from a mechanism of host immune evasion (34,35). Genetic exchanges, such as reassortment and recombination, lead to the generation of divergent virus progeny (36). Our previous studies identified reassortment and recombination of hantaviruses, including HTNV and MJNV, in nature (19,28,37). Using nearly complete sequences of SEOV S, M, and L segments, phylogenetic analyses demonstrated that S segments of group A SEOVs formed a cluster with those of group D SEOVs and that L and M segments of group A SEOVs showed a close phylogenetic relationship with those of group B SEOVs. The S segment of group C SEOVs grouped phylogenetically with group E SEOVs. However, L and M segments of group C SEOVs formed a distant genetic cluster from those of group E SEOVs. Phylogenetic analysis of SEOV S segments showed a differential pattern from that of SEOV M segments, indicating a genome organization compatible with genetic exchanges in nature. To clarify genetic events among SEOV worldwide, whole-genome sequences of the SEOV L segment need to be investigated. Application of multiplex PCR-based NGS will be useful in elucidating phylogenetic patterns of the SEOV L segment.
In conclusion, this epidemiologic survey of R. norvegicus rats in urban HFRS-endemic areas of South Korea identified the prevalence and distribution of SEOV. We applied multiplex PCR-based NGS to whole-genome sequencing of SEOV tripartite RNA from retrospective serum samples from HFRS patients and rat tissues. Phylogenetic analyses demonstrated the global distribution and genetic diversity of SEOV on the basis of nearly complete genome sequences. This study provides useful information for SEOV-based surveillance, disease risk assessment, and mitigation against hantavirus outbreaks.